Introduction

In a paper published in the 1990s, Janete Lins Azevedo calls attention to the theoretical-analytical debate in the scope of education policy, emphasizing the need for approaching education from a broader picture, proper of public policy that “represent the materiality of State intervention”, while at the same time, preserving the specificities of the education field. (^{Azevedo, 1997}, p. 6). The recommendation made by ^{Azevedo (1997)} over two decades ago remains not only current, but demands from the researcher an analytical basis capable of recognizing that the “state in action”, materialized by the policies produced by it, has been changing its outlines, assumptions, objectives and ways of providing services, in line with the deep political and socioeconomic changes occurred after the late 1970s in industrialized countries, and in the 1990s in capitalist-peripheral countries, such as Brazil (^{Pochman, 2007}). Changes of that order tend to cause significant structural change in the state and in its institutions, but they also affect the research field to the extent that one of its tasks is precisely to investigate the relations between the State and society, which requires reviewing analytical equations, so as to classify the new institutional arrangements and ways of producing educational policy, but also the ideological contributions that substantiate decisions of policy makers.

One of the most recent changes in the morphology of the State concerns the very logic assumed by it relative to forms of social solidarity, expressing trends of materialization thereof through - for example - entities of organized civil society, based on the argument they represent not only the overcoming of the State’s impasses, but also of those of the market, in the production of social policy such as education (^{Draibe, 1993}).

On that note, guidelines and tools structured in the context of New Public Management (NPM) passed to be one of the perspectives mobilized for handling the forms of restructuring of the State. Assuming that management by the private sector would be more efficient, more flexible and more accountable, several instruments used in this institutional environment were transferred to the public sector, creating in the states, more than in the Federal Government, the institutional conditions for “reinventing government” (^{Osborne; Gaebler, 1994}), whose duty should be, supposedly, “sailing instead of rowing”. The change to the meaning of governing, referring to new processes for regulating public education policy, associated to the use of new tools, but also in the interaction between public and private sector actors, for and not-for-profit (Rhodes,1996; ^{Kooiman, 2003}; ^{Tripodi; Sousa, 2018}), points to a complicated choice between the instrument of action and the resources that can be extracted from it.

Therefore, without the intention to summarize the “ecology” of instruments and tools mobilized by regulatory design assumed in Brazil, this article will be limited to analyzing “government contracting of results”, from the standpoint of the Agency Theory and Policy Instruments (IAP).

For addressing the question, the text is structured, in addition to this introduction, into another five sections and final considerations. The first section will present a brief definition of government contracting of results, with a succinct systematization of the debate held by education research. The method that set the framework for this article will be presented in the second section. Next, will be discussed the theoretical categories of Policy Instruments and their place in the State restructuring processes, and an inventory will be made of the main assumptions of the Agency Theory, for then, in a subsequent stage, examining the contractual design of the three states selected for the research: Goiás, Minas Gerais and São Paulo. The information of the states will be articulated with the technical-analytical perspectives assumed by the paper, followed by the conditions.

Government Contracting in the Public Sector

The government contracting approach, in the scope of the public administration, arises in the context of the fiscal crisis of the State, as a mechanism for supposedly improving the efficiency levels of the State machine, having been considered by authors such as ^{Pacheco (2004)} an important institutional innovation produced on a reformist level in the past few decades.

Conceptually, government contracting of results can be defined as an administrative instrument that structures a relationship negotiated between the various actors of the State machine and, eventually, between them and civil society (^{Saravia, 2005}). One of the arguments used in the defense of the use of contracts in the public sector argues that they would have the potential to improve the quality of services provided, through more flexibility, better performance and productivity, to the extent that is implemented, through them, an exchange of management autonomy for prior commitment to results.

In the country, the use of contracting dates back to the 1980s, and the Federal Railway System was the first to admit this form of management. In 1991, then-State Controller entity Vale do Rio Doce entered into an agreement called “performance agreement” with the Ministry of Infrastructure, whose contents was strongly tied to management contracts. In a previous stage, VALE itself negotiated these performance agreements with its subsidiaries.

In May of that same year, the Collor administration implemented the State Company Management Program through Decree 137, of May 27, 1991, for promoting efficiency and competitiveness of State companies, incorporating, to that effect, management contracts. From 1994 to 1996, Petrobrás also entered into a management contract with the Federal Government, through Decree 1.050, of January 28, 1994, whereby set conditions for government contracting. (^{Brazil, 1991}).

In 1998, after the creation of the Master Plan and State Reform (PDRAE), considering the New Public Management theoretical milestone in Brazil, Constitutional Amendment 19 translated government management of results in public administration as of the principles of efficiency, inserted into Article 37, § 8 of the Federal Constitution (§ 8) (^{Tripodi, 2014}).

In the field of education, one of the first Brazilian initiatives was the “State by Results”, implemented by the Aécio Neves Administration (2002-2010) and continued by Antônio Anastasia (2010-2014) in Minas Gerais, in the scope of the “Management Shock” Program. (^{Tripodi, 2014}).

Analyzing the program in Minas Gerais, ^{Augusto (2010)} argues that “political measures” that resulted in government contracting were “decided without the participation of the stakeholders”, in a linear and hierarchical manner, causing dissatisfaction by school inspectors in the State network, for not feeling heard and professionally recognized in the policy-making process (^{Augusto, 2010}, p. 11). In an article that compared the Minas Gerais education reform to the English model, in terms of adoption of management tools in education, ^{Prado (2012)} concludes that although there are common factors, Minas Gerais introduced, according to the author, only partially, results-based accountability, contrary to England that took this strategy all the way, establishing and divulging performance rankings by region, such as in the case of the performance league tables.

^{Segatto and Abrucio (2017)}, on their turn, upon examining four Brazilian states (Espírito Santo, Minas Gerais, Pernambuco and São Paulo) which established results-based management programs, conclude that although each state took its own reform trajectory, it was possible to identify in all of them a common root based on evaluations, curriculum and performance-based compensation.

What we can see is that a trend seems to be ongoing in the country, of implementing a “Contractual State”, although with particular design, pace and density, depending on each State’s ability, but also according to the political-administrative trajectories of each State. The findings of the Garcia conducted by ^{Garcia (2019)} points in that direction, by saying that it was possible to observe the existence of standardization relative to contracts in education management in 11 states and in Distrito Federal (Capital).

Method

The article is the partial result of a broader research, with a qualitative approach, that investigates the State borders in the field of Education, examining the conditions, outlines and privatization trends. There, two theoretical categories are presented and discussed – Policy Instruments and the Agency Theory - which seem to support the design of managerial education policy design, focused on government contracting of results. To that effect, the legal milestone that implements the reform in the States of Goiás, Minas Gerais and São Paulo between 2003 and 2014 are an intentional sample. It is important to point out that the analysis of the highlighted state propositions is outside of the scope of the article, in terms of the group of guidelines and outlines assumed, as well as investigation of implementation of these policies. We sought to investigate, in this paper, both the pertinence and the reach of both theoretical categories for examining managerial reforms. Selection of the states, as well as the time selection, is justified by literature (^{Abrucio; Gaetani, 2006}; ^{Tripodi, 2014}), which outlines the propagation of New Public Management (NPM) ideas as of 2003, when technicians who worked in the federal government, in the 1995 State Reform migrated to State governments and implemented managerial initiatives that were reformist in nature. Concerning procedure, the research is characterized as bibliographic and document-based, and the sources used were official publications, produced by state congresses and by the Executive Branch, in addition to academic papers on the theme. For examining the defended theory that education policy formulation and regulation logics, understood as tributary of NPM, there are theoretical-analytical tools that are coherent with their assumptions, assuming as work-dependent variables: i) creation of a structure of incentives in the field of education; ii) control of faculty absenteeism; and iii) dispersion of the education offer for Civil Society Organizations (OSCs); as for government contracting, it is a variable irrespective of analysis. From a theoretical standpoint, it is anchored on the Agency Theory and Policy Instruments approach.

Policy Instruments

The discussion around Policy Instruments (PI) is not new, after all, it can be found in classic authors such as ^{Max Weber (2009)}, in the form of the debate concerning the State and government ways of acting as domination techniques. To the author, the creation of bureaucracy, as a principle of rationality of societies assumes forms of exercise of power and, therefore, this is about a choice of instruments for governing. However, research on public policy tended to adopt, through the years, two stances relative to the instruments: on one side, the discussion was left aside, focusing the analysis on other aspects of policies and programs, such as actors, strategies or institutions. (Le Galès, 2012; ^{Vedung, 1989}). On the other side, we saw a trend as adopting them as a neutral, technical choice, though only in terms of a functional response, indifferently available (^{Lascoumes; Galès, 2012}). More recently, literature (^{Salamon, 2002}; ^{Hood, 1986}; ^{Howllet; Ramesh, 1993}; ^{Linders; Peters, 1989}) has been highlighting that, contrary to what was believed in the past, the selection of public action instruments are “deeply political choices”, that “[...] give some actors and, consequently their perspectives [of action], advantages in determining how policies will be enforced, [...] the choice of instruments helps determining which criterion will be used, and consequently, which interests will be the most favored as results” (^{Salamon, 2002}, p. 11, our translation).

In this sense, the PIs can be defined, according to ^{Salamon (2002)}, as an “identifiable method whereby collective action is structured for handling a public problem” (^{Salamon, 2002}, p. 19. Our translation). Instruments structure actions, i.e., the relations built by them are axiologically neutral, but they bear value and are “indissociable from the agents that program their use, that make they evolve and composed, based on them, communities of experts”. (^{Lascoume; Galès, 2012}, p. 22). And, precisely due to the fact they structure relations, they can be understood as “institutions” in the sense ascribed by neo-institutionalism, in a special way, as a rational choice, since they regulate standards of interaction between individuals or organizations, define the actors involved in the regulation of public programs and their roles.

Although it may seem that a large portion of instrument selection is made due to familiarity or a professional bias, ^{Linder and Peters (1989)} warn to the importance of knowing the expectations of decision-makers, or, in other words, what they expect when choosing one instrument in detriment to another (^{Linder; Peters, 1989}, p. 37, our translation). We must take into account, moreover, the fact that the PIs depend on a political context and social and cultural representation in a given society. That means that economic and fiscal factors have been conforming to new logics of thinking and of implementing policy, leading to transformation in government models, in the interaction between public and private sector players, in the pursuit for recomposition of the State structure, as well discussed by ^{Lascoumes and Galès (2012)}.

For example, the growth of privatization of social assets such as education, in most of the countries in the last decades, reflects, in the analysis of ^{Howllet and Ramesh (1993)}, “a fast and fundamental change in the standards of the policy instrument used” (^{Howllet and Ramesh, 1993}, p. 3, our translation). The authors refer to the use of management contracts, vouchers, public-private partnerships, among others, which passed to be mobilized based on new policy logics and management thereof. Therefore, one of the advantages of examining policy through its instrumentation is explained by the fact that the approach allows unveiling “the group of problems posed by the choice and use of the instruments (techniques, means of operating them, mechanisms) that allow materializing and operationalizing government action” (^{Lascoumes; Galès, 2012}, p. 20).

Agency Theory

The Agency Theory is born in the 1960s and 1970s, in the scope of the Transaction Costs School, within the New Institutional Economy (NIE). Despite the fact it came up in this field of knowledge, it has been taken over the lead role in the field of Sociology, especially in rational choice sociology, which, according to ^{Therét (2003)}, “it is just an extension of the new institutional economy of Political Science” [...] sharing “[...] a priori relative to institutions, the optimal calculation, assuming an instrumental-functionalist and government contracting-based position” (^{Therét, 2003}, p. 232).

According to ^{Jensen and Mecking (1976)}, formulators of the approach, the Agency relationship can be defined as a contract. In it, a principal hires an agent to carry out services to the benefit of the former, involving a delegation - at some level - of the decision-making authority to said agent. According to the authors, when individuals have a self-interest, both parties in the relationship will tend to maximize their own interests, whereby the agent will not act in the best interest of the principal, that hired it. Therefore, the principal would seek, in this theoretical perspective, to limit the divergences relative to the “disproportional” interest of the agent, by creating appropriate incentives, on which shall be levied, however, the cost of monitoring it (^{Jensen; Meckling, 1976}, p. 308, our translation). At the core of the Agency Theory debate are the supposed conflict of interest problems between the actors (individual and collective), transaction costs for reducing the asymmetry of information, the attempt to reduce opportunistic behaviors and the attempt to define more efficient contracts (^{Eisenhardt, 1989}).

The agency problem, therefore, lies in the risk the agent will act to its own benefit. Consequently, mechanisms are created, which theoretically, tend to reduce the possibility this actor acts in detriment of the principal, in a scenario of conflict of interest. There are, according to ^{Arrow (1985)}, two perspectives that substantiate the agency problem: the moral hazard and adverse selection. The first, the moral hazard, concerns information asymmetry between principal and agent. It would never have full information about the activities of the other, what it is doing, for example; the principal would lack capillarity to monitor the actions of its agent. As for adverse selection, it is related to the agent’s low ability to act on certain matters. In this case, the conflict between what the principal expects and what it is offered is due to the fact the agent does not know how to perform tasks under its responsibility (^{Eisenhardt, 1989}).

To ^{Eisenhardt (1989)}, the approach seeks to outline contractual frameworks or institutional proposals that align best and render the interests of both agent and principal compatible, such as using a monitoring and information system or formalizing results-based contracts.

The basic assumption of the approach is that individuals or organizations have self-interest, tending to maximize their own wellbeing, with a limited rationality and are risk-averse, the theory believes that a combination of inspection, incentives and punishment can solve the agency problem, allegedly causing the agent’s efforts to maximize the wellbeing of the principal.

By examining the design on the State, from a reformist perspective, using the Agency Theory as the reading key, ^{Przeworski (2006, p. 45-46)} argues that:

When some markets are missing and individuals have access to different information, are established between classes of actors agent x principal-type relations, governed by explicit and implicit contracts. Agents have certain information principals do not observe directly: agents know what motivates them, they have privileged knowledge of their abilities, and they may have the chance to observe things the principles cannot. There are some actions which, at least in part, are carried out without the knowledge of the principal. In generic terms, therefore, the problem the principal has to face is this: how to induce the agent to act in its interest (that of the principal) [...] Performance of companies, governments and of the economy as a whole depend on the design of institutions that regulate these relations. What matters is whether those employed have incentives to maximize their efforts [...].

The author, for illustrating the concept, resorts to the analogy of “going to the mechanic” because the care is making strange noises. The mechanic, representing the agent, in this case, solves the problem and claims to have changed the shock absorbers and that it required five hours of labor. The car owner, the principal, according to ^{Przeworski (2006, p. 45)}, “[...] has to pay the mechanic the sufficient for him [the mechanic/agent] to want you [the car owner/principal] to go back to him at a later time, but has, moreover, to find a way for him [the mechanic] to know you [the principal] will only come back if he did a good job”.

It is imperative to note that for this approach, rewarding or sanctioning, as means of creating an incentive structure, is not sufficient. It is especially important that the results of the contract, whether implicit or explicit, are “publicized”, whereby the “loss of a protected future due to loss of good reputation will be higher than the gain obtained through fraud or through a poorly done job”. As for the case of “common workers, who do not have a reputation the employer can search without incurring high cost, literature recommends probation stages and opportunity for promotion based on good performance” (^{Monsma, 2000}, p. 89).

The type of incentive determined by/in contracts, was one of the aspects considered by ^{Eisenhardt (1989)} as a subject deserving special attention. To the author, it is important to be clear that some results are more complex to measured or observed than others, since we cannot lose sight that according to the type of contract, its framework, assumptions and structuring are tributaries of scenarios where the principal considers there is full or imperfect information.

Government Contracting in States: characteristics and reach

In this section, the designs, the logics of contractualization in the educational scope will be presented, based on the experience of three Brazilian states: Goiás, Minas Gerais and São Paulo.

Government Contracting in Education: theoretical articulations

The three states examined passed to implement, in a more systematic manner, as of 2003, reforms based on NPM principles. Although transmitting to each of them a different pace and density to the reform process, it is possible recognize in all of them the conceptions of State and of management disseminated by the theoretical framework of the federal State reform, of 1995, in the first term of President Fernando Henrique Cardoso (1995-1998), with especial emphasis on the government contracting of results.

Therefore, in Minas Gerais that prioritized management contracts or partnership agreements structured on predetermined goals by school unit, focused on the student performance in SIMAVE, and formally signed between the Department of Education and the schools. As for Goiás, the contracts sought to formalize the relations between the State and non-state actors, especially SOs, in offering High School enrollments. São Paulo, on its turn, prioritized diffuse contracts, implicit in nature, without formalizing execution between education professionals and the central agency, although there was contracting of goals relative to management, student performance in IDESP, faculty attendance, as well as participation in continued formation programs offered by SEE/SP and bonuses. This contractual design can be understood as symbolic, to the extent if had not been expressed in formal contractual provisions and which, for that reason, did not impose sanctions in case the goals were not met, although maintaining the rewards.

If the reform designs are different, the instruments, on their turn, move in the same direction, and also seem to follow the assumptions that substantiate their choice. First, we must pay attention to the fact that government contracting is an instrument of “public action” and not of public policy. The former, contrary to the latter, is enforced by actors outside the State’s purview. Therefore, contracts, when structuring arrangements with actors outside the State, pass to have the potential to influence and change the functioning and the very structure of the State. Not by chance, Goiás government, in Decision 596/2015, when justifying the reasons for using government contracting with SOs, in the offer for High School Education, clarifies that it is “[...] evidently a managerial reform, since it allows reconstructing the State in new basis, recognizing that the perspective of public policy goes beyond the government policy perspective, in which context arises and important and undeniable participation of the Third Sector” (^{Goiás, 2015b}, p. 2).

In this sense, “looking” to the selection of the instrument allows identifying new characteristics and logics that have been imposed in the field of policy and which expresses deeper social welfare State restructuring in the country, because the choice of the instrument is closely related to the assumptions articulated by it. Therefore, in this article, we defend that the news ways of conceiving and conducting education policy demand mobilization that new theoretical-analytical input proving the premises on which these policies were designed.

Therefore, it is important to highlight that government contracting is, fundamentally, one of the NPM tools, whose principles are tributaries of Business Administration, but predominantly, of New Institutional Economics. The main ideas that substantiate this school of thought are organized around the introduction of market logics in the scope of public administration, based on i) the creation of incentive structures, ii) emphasis on competition encouraged by the use of contracts and by the quasi-market, iii) client choice theory; and iv) division of state bureaucracy between policy designers and enforcers (^{Hood, 1991}; ^{Rhodes, 1996}). From the theoretical standpoint, one of its main contributions is the Agency Theory. In this perspective, questioning the use of contracts at the level of the State requires a double analytical movement of identifying in the principles of the NIE, especially in the approach of the Agency Theory, the reasons and substantiation for the choice, at the same time as examining the compatibility (or lack thereof) of its uses in the field of education.

The first question pose is, consequently, knowing if an approach whose basic premise is aligning conflicting interests has the potential to dealing with public administration, and especially with education policy. Would it be reasonable to conceive the existence of a conflict of interest relative to education professionals and the government, motivated by different preferences in discharge of the office? Which would be the opportunistic behaviors that contracts would tend to limit, on an education level, where the actors are public servants? Who would be the actors willing to maximize their own private interests?

The base that sustains the choice of government contracting as means of regulating part of an economic view for maximizing gain, is assuming the existence of limited rationality in transactions and understanding that there is information asymmetry between the parties, and that, at some point, one party will tend to benefit from it, through opportunistic behavior. Therefore, the use of contracts needs to be read from this perspective in the case of education, since its is inserted into a managerial perspective of administration.

The choice for using contracts suggests, in theoretical-analytical terms, that the relationship between the Executive Branch and education professionals is structured from a point of initial mistrust, by the government, in the ability of the public servant to meet its duties, especially relative to student learning and teaching. According to the approach, the problem with the agency would be, then, in the fact that the central education agency lack sufficient capillarity to reach school units and particularly its faculty, for monitoring and/or controlling its behavior, according to what would be happening in school units and being taught in classrooms. Therefore, working with the scenario that schools would not be, for example, implementing curricula implemented by the central agency, which would be, hypothetically, observed as a result of large-scale evaluations, the principal, who in this case is the central administration, structures a targets system through contracts with schools so agents (especially teachers) will thereby work in alignment with the interests of the principal. I.e., it expected that teachers will pass to teach the predetermined curriculum, established in the form of results-based targets and formalized by the execution of a management contract.

Since this is an economic theory that recognizes the individual, a priori, as self-interested, the instrument of public action, namely, the contract, needs to have sufficient power to compel said agent to control its supposed opportunistic behavior and work to maximize the interests of the principal. From that perspective, the contract passes to have its foundation on an incentives-based structure, which provides for rewards and sanctions, for chancing conflicting behaviors. As mentioned before, for the Agency Theory, creating incentives is not sufficient, it is important that the results of the “transaction” are publicized, whereby the quality of the information disclosed has the potential to impact future behavior. To that effect, the choice for the format to disclose large-scale student evaluation results and education indices such as IDEB, mostly done through comparisons and ranking, has played a strategic role in the so-called agency problem, since the very reasons presented by the Goiás Governor for using the SOs in education, are primarily justified by IDEB.

We note that the government contracting instrument, from the Agency standpoint, seek to address both the moral hazard and adverse selection in the field of education. To the extent that contracts, implicit or not, set goals relative to student proficiency in evaluations, but also based on faculty absenteeism, participation in formation courses and finally, relations between the State and OSC’s, we see that the “mistrust” relative to the education field is limited not only to the scenario where professionals do not do what they are supposed to do, but also that they do not know how to do what must be done.

One of the questions that must be debated and seems to be a great contradiction of the approach applied to the public sector, is the (non)existence of diverging interests at stake, since, in the case of public service, the State and its education professionals tend to defend the same interests, namely, a successful learning process and effective management of schools; after all, as their title itself suggests, they are “servants” of the Public (State). If the answer to that question is built in terms according to which these professionals do not know what to do or how to teach or how to manage and that, consequently, the problem that the agency seeks to correct is adverse selection, the choice for government contracting, seems to have the potential to encourage undesirable practices, already examined by research in the field (^{Capocchi, 2017}; ^{Cerdeira, 2018}), such as preparation and selection of students for tests, for meeting contract targets.

The nature of the legal regime to which public education servants are linked is another element that allows unveiling the intentionality that underly the choice of the instrument such as contracts and the stressing of its purported neutrality. That is due to the fact that, according to the nature of the statutory regime, it also is, necessarily so, “unilateral”, based on the “legal provision named statute, characterized by submission of the servant to this law [...]” (^{Anastasia, 1990}, p. 76). In this sense, the “public servant” waives, upon being invested into office “the autonomy of free workers as to setting working conditions” and accept “administrative regulation, decreed on a unilateral basis” (^{Romita, 1975}, p. 14). For that very reason, the State, as the lawful holder of the means necessary for investing administrative resources (^{Weber, 2009}) has the prerogative to enforce that is determined as a duty of the public servant, and therefore, there is no reason for executing a contract that will compel public servants to properly discharge their duties. Consequently, bearing this legal rule in mind, it is reasonable to conclude there is a need to note that the ideological input underlying government contracting in State public education; and, in this sense, the experience in Goiás is pedagogic. In Decision 596/2015, the governor states that “the institutional model of social organizations present clear advantages relative to state organization, which is why i refer to the case of acquisition of goods and prices and contracting of services”. According to Perillo that is due to the fact that “because management contracts encourage efficient procedures, adequate to public needs, contributing to diminish rigidity elements that characterize bureaucratic cultures [...]” (^{Goiás, 2015b}, p. 2).

Examining the justifications expressed in Government Decision, such as in Article 7-A, art. 9 and sole paragraph, Article 14-A, Article 14 B and §7, of State Law 15.503/2005, which approves SOs in Goiás, what we see is the possibility that all resources considered inefficient by the governor, on a state scope, can be assigned to Social Organizations. That is proved by the possibility, allowed by Law, that public servants will be detailed to SOs paid by the State. Are regulated, moreover, i) the perspective of resources being transferred to SOs, as investments, in the beginning or during performance of the contract, ii) expansion of the existing physical infrastructures, iii) acquisition of necessary complementary chattel; and iv) use of assets, facilities and public equipment necessary for meeting the contracted goals/targets (^{Goiás, 2005}).

Considering this scenario, it interesting to note that flexibility, held as a structuring part of the SOs, and considered by the governor as a decisive element for improving education, is more closely related to investments in education made by the State to the benefit of Social Organizations than a “clear advantage” per se, of the contractual model, as some would have you believe. The question posed, consequently, is: if the public servants can be the same, if the infrastructure can be the same, if many OSC’s can be solely maintained with State resources and, moreover, receive State investment, what is the justification for that undertaking? Everything points to the fact that the choice made according to these logics of offer and regulation are less based on evidence of what the State is able to do, and more in what, ideologically, many seek to convince the majority that the State is not able to do.

Conclusions

PDRAE, considered a milestone of management reform in the 1990s, in the scope of MARE, played an important part in the configuration of reforms on a state level, based on NPM principles, especially after 2003. Although the depth and the reach of reforms on an education level have been distinct, government contracting of results was one of the instruments prioritized by the States of Goiás, Minas Gerais and São Paulo. In the case of the first two states, contracts were used, respectively, for aligning supposed diverging interests that formulate and implement education policy, all limited to the state’s purview, and for establishing relations between the State and the SOs in the offer for High School education. In the case of São Paulo, contracts can be defined as implicit, to the extent that the goals agreed between the actors, also State-controlled, were produced by government acts and not specifically through management contracts. The article provides evidence that education policies limited to the New Public Management logics and, therefore, adopting NIE principles, demand a theoretical-analytical input accounting for assumptions and beliefs involved in the matter. In this sense, the study points to the fact that the Agency Theory and the Policy Instruments approach provide the input necessary for examining these proposals, considering that its main premise is based on the mistrust by education policy makers relative to the ability of policy executors in enforcing them, whether for – allegedly – not wishing to do so, or for not knowing how to proceed. From that theoretical perspective, the work of empirical reconstitution allows us to say that the use of results of student performance evaluations, on a large scale, became a prioritized instrument of public action on which government contracting has been based, for the purpose of aligning supposedly diverging behaviors. The article, finally, offers elements to say that everything points to the fact that the choice for these instruments and approaches by the government is substantiated by an ideological input that recognizes, a priori, in the State, a certain inability to offer and regulate education policy.